First up, just let me say I couldn't be more grateful for all the info I've found lurking these forums (unregistered) that I've found over the past month or so. So, thank you to all the great people on here willing to share their knowledge.
On to the topic at hand. I love bass 😛 and over the years have been getting more interested in the topic, specifically when it comes to horn loaded subwoofers (and mid/bass and highs of course!). I bought a polk audio psw505 a couple years back, and while it does get loud, it also sounds messy. Just not very clean at all. So, being on a bit of a small budget, and not willing to spend money on cheaper, larger woofers, I decided it made sense to design a tapped horn around some of the smaller drivers available. For this project I used the tang band 1138smf, a 5.25in sub driver (which I'm sure you're all familiar with) with some pretty fantastic results.
Designed with hornresp. Which i have a question about below. Decided to go with steps rather than a pure tapered design for ease of construction (am a novice at woodworking)
Tang Band tapped horn project - Album on Imgur
I dont have professional sound level metering equipment, or even a decent microphone... but I do have an app on my smartphone 🙄 I measured +/- 3db from 25-60hz, a large (+6) spike from 60-80, then it leveled out to 100hz before dropping off considerably. I sincerely doubt the levels my phone reported (98ish db, @ 1w/1m) so I'm not bothering to upload any pictures yet, but I just wanted to see how smooth the response actually was considering my room is very mode-y.
On to the question(s):
Is there any way to smooth out the peak from 60-80? Or is that a consequence of the box design?
In my hornresp sims, before I built the actual box, I designed it to be relatively flat from 80 to 35hz in 2pi space. I know it's not a perfect simulation, however I was not expecting the extra 10hz below the design. Now, re-measuring the box itself, I'm showing a tuning of 40-100hz. Yet it hits quite a lot lower than this. Why is that? I'm not complaining, just striving to understand why that is the case. I know the room has a lot to do with how subwoofers behave in general, and this TH is sitting under my desk in a smallish nook in an L shaped room.
I must say I'm blown away by the sensitivity of this design. I can only assume a front loaded horn (if I ever have the space for one) would be even better in that regard. Also, how clean and clear these frequencies are being reproduced astounds me. It nearly keeps up with the polk sub, which is around 430w rms. A 40w sub with approximately 20 watts into it is nearly as loud as it, all thanks to box design. I'm giddy, lol.
Anyways, thanks for taking a look at my project. I can post hornresp inputs when I'm back at my computer in a couple of days, if theres any interest.
P.s. a coworker has said he has the appropriate gear to take real measurements, no idea when he will bring it to work with him though.
On to the topic at hand. I love bass 😛 and over the years have been getting more interested in the topic, specifically when it comes to horn loaded subwoofers (and mid/bass and highs of course!). I bought a polk audio psw505 a couple years back, and while it does get loud, it also sounds messy. Just not very clean at all. So, being on a bit of a small budget, and not willing to spend money on cheaper, larger woofers, I decided it made sense to design a tapped horn around some of the smaller drivers available. For this project I used the tang band 1138smf, a 5.25in sub driver (which I'm sure you're all familiar with) with some pretty fantastic results.
Designed with hornresp. Which i have a question about below. Decided to go with steps rather than a pure tapered design for ease of construction (am a novice at woodworking)
Tang Band tapped horn project - Album on Imgur
I dont have professional sound level metering equipment, or even a decent microphone... but I do have an app on my smartphone 🙄 I measured +/- 3db from 25-60hz, a large (+6) spike from 60-80, then it leveled out to 100hz before dropping off considerably. I sincerely doubt the levels my phone reported (98ish db, @ 1w/1m) so I'm not bothering to upload any pictures yet, but I just wanted to see how smooth the response actually was considering my room is very mode-y.
On to the question(s):
Is there any way to smooth out the peak from 60-80? Or is that a consequence of the box design?
In my hornresp sims, before I built the actual box, I designed it to be relatively flat from 80 to 35hz in 2pi space. I know it's not a perfect simulation, however I was not expecting the extra 10hz below the design. Now, re-measuring the box itself, I'm showing a tuning of 40-100hz. Yet it hits quite a lot lower than this. Why is that? I'm not complaining, just striving to understand why that is the case. I know the room has a lot to do with how subwoofers behave in general, and this TH is sitting under my desk in a smallish nook in an L shaped room.
I must say I'm blown away by the sensitivity of this design. I can only assume a front loaded horn (if I ever have the space for one) would be even better in that regard. Also, how clean and clear these frequencies are being reproduced astounds me. It nearly keeps up with the polk sub, which is around 430w rms. A 40w sub with approximately 20 watts into it is nearly as loud as it, all thanks to box design. I'm giddy, lol.
Anyways, thanks for taking a look at my project. I can post hornresp inputs when I'm back at my computer in a couple of days, if theres any interest.
P.s. a coworker has said he has the appropriate gear to take real measurements, no idea when he will bring it to work with him though.
Last edited:
If your TH performs very different to what you sim'd, there's only two real reasons I can think of for that happening:
1. You didn't build what you sim'd
- How did you account for that damped chamber at the the start of the TH?
- Measure the length again to see if you got it correct
- Measure the impedance curve and determine Fb from that.
2. The driver's specs in the sim don't match the actual specs of the driver you're using
- Being a Tang-Band driver, I'd have measured its parameters first before doing a sim
1. You didn't build what you sim'd
- How did you account for that damped chamber at the the start of the TH?
- Measure the length again to see if you got it correct
- Measure the impedance curve and determine Fb from that.
2. The driver's specs in the sim don't match the actual specs of the driver you're using
- Being a Tang-Band driver, I'd have measured its parameters first before doing a sim
what's up with the dead space on the front of the enclosure, put a thin panel there and you've got a dejambe. i could be wrong but that's likely where the peak is coming from.
all tapped horns require a high pass filter because there's no excursion control/damping at low frequencies.
all tapped horns require a high pass filter because there's no excursion control/damping at low frequencies.
Last edited:
I never realized Tb drivers could be so far out of spec. I'll have another look at measuring t/s parameters and see what I get. Is this pretty common with them in guessing?
I couldn't accohhhunt for the damped chamber. In my head the damping was going to change, this was really just an experiment for me. The right side panel is removable so as to experiment with different damping/materials etc, and is sealed with rope caulk 😛
And the dead space will soon house the mid and high drivers making this a full range setup. I didn't think having it there as it is, considering the bracing internally, would be as much of an issue as it is, but you may be right. Also, I'm more of a cajon guy myself, I've gotten too many blood blisters playing djembes 😉
Thanks for the input guys. I realize this is a much more haphazard approach than people here may be used to... but I tend to learn more this way, rather than just reading until my eyes bleed.
I don't understand how such a small mouth can be horn loading bass frequencies. Can you show us some hornresp screenshots?
Wanted to add for clarity: the graph is in 1pi space even though the inputs state 2pi. 1pi is essentially how it is placed at the moment.
The horn segments should probably be PAR (for Parabolic), rather than Con (for Conical) or Exp (for Exponential), as that's usually the best approximation for a horn folded into a rectangular box.
The big spike that you measured at 60~80 Hz does show up in the sim though.
FWIW, I ran a sim of that driver in a simple offset driver "pipe" (50 cm2 csa, 230 cm long, driver at 74 cm, stuffed a bit up to where the driver is located), and the results look pretty decent, with usable response up to over 100 Hz, and net box size is just over 11.5 liters. The driver will likely have to mounted facing into the box though 🙂. It's not something I would build anyway until I measured the driver's parameters.
A consequence of a relatively high Qes and too large a pipe for the tuning.
The distance between the driver's front back = ~34400/4/320 = ~26.9 Hz base tuning and a ~352 cm/~24 Hz pipe cutoff [Fp], but due to its 'weak' motor ideally needs to be tuned lower for flattest response to get the large net Vb required, otherwise need to stuff the TH same as any other TL.
GM
Thank you for the explanation. That makes things a bit more clear, I'm going to remove the side panel at work today and add stuffing to the line and see where that gets me as it looked promising in HR.
Quickly realized through exploring all the options in HR that this design only allows about 10w thriugh the driver before reaching xmax near 40hz, which corresponds with what I've noted in the real world. Still plenty loud, just a bit disappointing. Live and learn I suppose.
Yeah I'm thinking that will be my next step, get actual measurements for the driver vs. what PE says they are.
I've also found the spike isnt quite as pronounced (through listening) as the graph indicates, and that it actually adds a bit of musicality to the output since that's right about where kick drums usually play, giving it some nice punch without having to touch an eq.
I may also try out the pipe design, as we have pretty much every size of sonotube available at work and are encouraged to experiment with speakers (within reason). Will definitely respond here if that happens.
Thank you guys for chiming in, I simultaneously feel I'm beginning to understand what makes TH work and also that i have no idea what I'm doing (which is closer to the truth i think). I really appreciate it!
- Status
- Not open for further replies.